A healthy and productive online community can be a tremendous force for good. A bad one however can easily turn poisenous. This post will focus on some of the warning signs that will emerge when your community is losing its edge. It assumes that you have a clearly defined purpose for your community against which to judge its effectiveness. If your community doesn't yet have a purpose, go and get one – now.
Ok, those 5 tell tale signs, as promised.
5 signs your community is going bad
- Discussions are losing their focus – whilst a little off topic chat is fine as it helps to bond the community together, you need to remember why your community exists and keep it focused on that goal. If discussions are drifting to a related topic it might be time to create a seperate community for that purpose, if it's off topic completely however you need to drag it back on track.
- Your new topic/replies ratio is bad – if you're getting a lot of topics and no replies it suggests a disengaged community. Likewise your discussions need to be moving your community towards its purpose.
- The tone is turning ugly – you want your community to be a respectful place with wide engagement. If a minority dominate discussions and a clique has emerged then it is not a good community for wide and varied discussions. You want dissent, but you need it to be respectful dissent.
- New members aren't sticking – ask yourself how the makeup of your community is developing. Are new members transitioning into respected members or do they fail to stick around after their initial activity? This is a sign that the community isn't attractive to those outside the established core.
- You stop making progress – the point of your community is to make progress towards its purpose. If it's failing to do this then it isn't fulfilling its role.
These are a few relatively simple things you can do on a regular basis to check the temperature of your community to ensure it is still doing the things you want it to.
What other signs are there that your community is going in the wrong direction?
They probably cover most things. Will you do a post on how to get things back on track should any of these occur?
Yes I probably should 🙂 Leave it with me. Thanks Paul.
I think this is related to your metric of the "new posts/replies ratio", but I have also found it useful to look at: participation rate (% of site visitors that add content in any way), visit density (% of members of the community that visit the site within a given time period, e.g. a week), and average time per visit. If you can track these measures over time, I think you can get a very good notion of the health of the community. For example, I find that visit density often gives an indication of whether there is sufficient new content on the site that is attractive to members. So, if it drops off you can look into the amount of new content and also the topics being covered by that new content. Changes in these metrics can also be used to motivate more direct action (e.g. talking to members about changes they have perceived in the community, about why their participation has changed, etc).
Thanks Terry (and welcome to the blog). I think that's a good metric. I used to find with my old vBulletin community that the community was at its healthiest when the activity was spread over a wide number of people. It wasn't so good when a few were making all of the posts. With the CMI community however it wasn't easy to track that as Drupal didn't provide the stats.
The general web stats however are there in any web analytics package so should be easy to come by and, as you say, provide you with more info about how people are finding your community.
A question for you. How would you gauge success in a LinkedIn group (for instance) where stats are a bit harder to come by?
Hi Adi – I haven't managed a LinekdIn group before, so I haven't given it much thought. It does look like the Growth and Activity sub-tabs in the analytics area provide pretty good data on content generation over time. Combined with total group membership, that gives a fairly good picture of active participation. However, it doesn't look like LinkedIn provides any data on more passive participation (# of articles read, avg. visit times, etc). Maybe the folks at LinkedIn are listening 🙂
Yes they have expanded their group analytics nicely, and hopefully will continue to do so. I'd like to advocate a slightly softer, more qualitative approach to measurement however that revolves around the purpose of your community. For instance not all posts are of equal value. Some will be off topic water cooler type discussions, which whilst useful are not why your community exists.
So rather than counting all posts I'd personally go for counting just the content that is around your communities purpose. So lets say your community exists to provide customer support, you could measure a) the number of support questions posted, b) the number successfully answered, and c) the number not successfully resolved. That way you can easily get your community back into business focused ROI by comparing for instance a successful support story with the cost of a similar success via your call centre.
Check this post out http://adigaskell.org/2011/12/13/how-to-…
It talks about social ROI and business ROI. The social is the number of successful stories you get from your community, and the business is how those wins earn you/save you money.
I think that metrics (in the sense of automatically measured quantities) are always just a starting point. So, I agree with you that you will get more useful information from your data when you apply the "human touch" — for example, by classifying the nature of postings as you suggest. In a similar vein, I think that analytics are often useful simply as drivers of further information gathering; that is, helping you focus on what you should ask your members, or perhaps suggesting alternative strategies to a/b test.
It can be a bit of a poisened chalice. Those that don't really get social media are often swayed by the large numbers available, without really giving thought as to what those numbers mean. There are still so many people that chase incredibly high Twitter followerships for instance without any other means of measuring whether that's a good thing or not.
From your comment in the other article about community strategy though it sounds like you have very much the right approach to things.